Ryerson University GRASCan 2015 GRASCan 2012 Ryerson University 2 Emotions are contagious Graph burning Anthony Bonato 3 KramerGuilloryHancock14 study of emotional or social contagion ID: 759109
Download Presentation The PPT/PDF document "1 How to burn a graph Anthony Bonato" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
1
How to burn a graph
Anthony BonatoRyerson University
GRASCan 2015
Slide2GRASCan 2012, Ryerson University
2
Slide3Emotions are contagious
Graph burning - Anthony Bonato
3
(
Kramer,Guillory,Hancock,14):
study of
emotional
or
social contagion
in Facebook
t
he underlying
network is an essential
factor
in-person interaction and
nonverbal cues are not necessary for the spread of the
contagion
Slide4Graph burning - Anthony Bonato
4
Slide5Modelling social influence
general framework:nodes are active or inactiveactive nodes are introduced and influence the activity of their neighbours
Graph burning - Anthony Bonato
5
Slide6Models
various models:(Kempe, J. Kleinberg, E. Tardos,03)competitive diffusion (Alon, et al, 2010)literature in graph theory:dominationfirefightingpercolation
Graph burning - Anthony Bonato
6
Slide7Memes
memes: an idea, behavior, or style that spreads from person to person within a culture
Graph burning - Anthony Bonato
7
Slide8Meme theory explained
Graph burning - Anthony Bonato
8
Slide9Quantifying meme outbreaks
meme breaks out at a node, then spreads to its neighbors over timememe also breaks out at other nodes over discrete time-stepshow long does it take for all nodes to receive the meme in the network?
Graph burning - Anthony Bonato
9
Slide10Burning number
G a connected, simple graph there are discrete time-steps or roundseach node is either burning or non-burningif a node is burning, then it remains in that state every round, choose an additional non-burning node to burnonce a node is burning in round t, in round t + 1, each of its non-burning neighbors becomes burningchosen nodes: activatorsprocess ends when all nodes are burning the burning number of a graph G, written by b(G), is the minimum number of rounds for all nodes to be burningwell-defined, as bounded above by |V(G)| (even (G)+1)
Graph burning - Anthony Bonato
10
Slide11Example: cliques
b(Kn) = 2
Graph burning - Anthony Bonato
11
Slide12Paths
burning sequence: (v3,v7,v9)sequence of activatorsTheorem (Bonato,Janssen,Roshanbin,14)b(Pn) =
Graph burning - Anthony Bonato
12
1
2
3
2
2
3
3
3
3
v
1
v
2
v
3
v
4
v
5
v
6
v
7
v
8
v
9
Slide13Proof of lower bound
suppose (x1,…,xk) is a burning sequence for Pnthen: Nk-1[x1] Nk-2[x2] N0[xk] = V(G) (1)as |Ni(x)| ≤ 2i for all nodes x, we have by (1) that: + k = 2k(k-1)/2 + k = k2 ≥ n
Graph burning - Anthony Bonato
13
Slide14Trees
rooted tree partition of G:
Graph burning - Anthony Bonato
14
collection
of rooted trees which are subgraphsof G, with the property that the node sets of the trees partition V(G)
x
1
, x
2
, x
3
are activators
Slide15Trees
Theorem (BJR,14)b(G) ≤ k iff there is a rooted tree partition with trees T1,T2,…,Tk of height at most k-1, k-2, …,0 (respectively)such that for all i, j, the roots of Ti and Tj are distance at least |i-j|.
Graph burning - Anthony Bonato
15
Slide16Trees
note: if H is a spanning subgraph of G, then b(G) ≤ b(H)a burning sequence for H is also one for GCorollary (BJR,14)b(G) = min{b(T): T is a spanning tree of G}
Graph burning - Anthony Bonato
16
Slide17Bounds
Corollary (BJR,14)b(Cn) =If G has a Hamiltonian path, then b(G) ≤ burning is not monotonic in general on subgraphs; even for isometric subgraphseg b(C5) = 3 > b(W5) = 2Lemma (BJR,14) If H is an isometric subtree of G, then b(H) ≤ b(G).
Graph burning - Anthony Bonato
17
Slide18Aside: spider graphs
SP(3,5):Lemma (BJR,14) b(SP(s,r)) = r+1.
Graph burning - Anthony Bonato
18
Slide19Bounds
Theorem (BJR,14)If G has diameter d and radius r, then≤ b(G) ≤ r+1.tight: upper bound: spider graphslower bound: paths
Graph burning - Anthony Bonato
19
Slide20Coverings
Theorem (BJR,14)If C1,C2,…,Ct cover G, and each Ci is connected of radius at most k, then b(G) ≤ t + k.(G): k-distance domination numberCorollary (BJR,14) k}
Graph burning - Anthony Bonato
20
Slide21How large can the burning number be?
Conjecture (BJR,14): b(G) ≤ .by using corollary on we have that: b(G) ≤ 2-1.(Coudert,Nisse,Roshanbin,15+): b(G) ≤
Graph burning - Anthony Bonato
21
Slide22Nordhaus-Gaddum type results
Theorem (BJR,15+)If n ≥ 2, then 4 ≤ b(G) + b() ≤ n+2. If G is connected and n ≥ 6, then b(G) + b() ≤ -1.Theorem (BJR,15+)If n ≥ 6, then 4 ≤ b(G)b() ≤ 2n. If G is connected, then b(G)b() ≤ n + 6.Conjecture (BJR,15+): b(G)b() ≤ n + 4.
Graph burning - Anthony Bonato
22
Slide23Graph burning - Anthony Bonato
23
Iterated Local Transitivity (ILT) model(Bonato, Hadi, Horn, Prałat, Wang, 08)
key paradigm is
transitivity
: friends of friends are more likely friends; eg
(Girvan and Newman, 03)
iterative cloning of closed neighbour sets
deterministic
local
: nodes often only have local
influence
evolves
over
time
, but retains
memory of initial graph
Slide24Graph burning - Anthony Bonato
24
ILT model
begin with a graph
G = G
0
to form the graph
G
t+1
for each vertex
x
from time
t
, add a vertex
x’
, the
clone of
x
, so that
xx’
is an edge, and
x’
is joined to each neighbor of
x
order of
G
t
is
2
t
n
0
Slide25Graph burning - Anthony Bonato
25
G
0 = C4
Slide26Graph burning - Anthony Bonato
26
Properties of ILT model
average degree
increasing
to
∞
with time
average distance
bounded by constant and converging
, and in many cases
decreasing
with time; diameter does not change
clustering
higher
than in a random generated graph with same average degree
bad expansion
: small gaps between 1
st
and 2
nd
eigenvalues in adjacency and normalized Laplacian matrices of
G
t
Slide27Burning ILT
although ILT generates graphs with exponential order/size, the burning number is constant:Theorem (BJR,14) For all t, b(Gt) {b(G0), b(G0)+1}.
Graph burning - Anthony Bonato
27
Slide28Cartesian grids
Graph burning - Anthony Bonato
28
Slide29Cartesian grids
Theorem (BJR,14)If 1 ≤ m ≤ n, and G is the m x n Cartesian grid, then we have the following:If m = O(), then b(G) = If m = ), then b(G) =
Graph burning - Anthony Bonato
29
Slide30Sketch of proof
consider upper bound in the case m = O() idea: using a covering by t closed balls of radius r (diamonds), with r to be determinedgives upper bound for b(G) of t+r by covering theorem
Graph burning - Anthony Bonato
30
2r+1
2r+1
Slide31Sketch of proof
now let r = (Mitsche,Prałat,Roshanbin,15+) derived constantsfor the n x n grid, b(G) =
Graph burning - Anthony Bonato
31
Slide32Complexity
Burning number problem:Instance: A graph G and an integer k ≥ 2.Question: Is b(G) ≤ k?in NP
Graph burning - Anthony Bonato
32
Slide33Burning a graph is hard
Theorem (BJR,14+) The Burning number problem is NP-hard.Further, it is NP-hard when restricted to any one of the following graph classes:planar graphs disconnected graphsbipartite graphsreduction from planar 3-SAT
Graph burning - Anthony Bonato
33
Slide34Gadgets
Graph burning - Anthony Bonato
34
Slide35Burning a graph is hard
Theorem (BJR,15+) The Burning number problem is NP-hard when restricted to trees of maximum degree 3.reduction from a certain subset-sum problem
Graph burning - Anthony Bonato
35
Slide36Random burning
select activators at randomwe consider uniform choice with replacement
Graph burning - Anthony Bonato
36
Slide37Cost of drunkeness
bR(G): random variable associated with the first time all vertices of G are burningb(G) ≤ bR(G)C(G) = bR(G)/b(G): cost of drunkenness
Graph burning - Anthony Bonato
37
Slide38Drunkeness on paths
Theorem (Mitsche,Pralat,Roshanbin,15+)C(Pn) = first and second moment methods
Graph burning - Anthony Bonato
38
Slide39Other random burning models
choose activatorswithout replacementfrom non-burning verticesfor (1), cost of drunkenness on paths is unchanged, asymptoticallyfor (2), cost of drunkenness is constant
Graph burning - Anthony Bonato
39
Slide40Future directions
conjecture: b(G) ≤ burning in grids strong, hexagonal, triangulargrids?3-dimensional?infinite grids?burning in graph products Cartesian, strong, categorical
40
Graph burning - Anthony Bonato
Slide41Future directions
random graphs and cost of drunkennessbinomial, geometric random graphs (MPR,15+)random regular?drunkenness in hypercubes?graph bootstrap percolationvertices burn if joined to r >1 burning verticesburning in models for complex networkspreferential attachment, copying, geometric models?
Graph burning - Anthony Bonato
41
Slide42Graph burning - Anthony Bonato
42